Guidelines for the first line management of classical Hodgkin lymphoma

Transcription

1 guideline Guidelines for the first line management of classical Hodgkin lymphoma George A. Follows, 1 Kirit M. Ardeshna, 2 Sally F. Barrington, 3 Dominic J. Culligan, 4 Peter J. Hoskin, 5 David Linch, 2,6 Shalal Sadullah, 7 Michael V. Williams 8 and Jennifer Z. Wimperis 9 for the British Committee for Standards in Haematology 1 Department of Haematology, Addenbrookes Hospital, Cambridge University Teaching Hospitals, Cambridge, 2 Department of Haematology, University College Hospital London Hospitals NHS Trust, 3 St Thomas Division of Imaging, PET Imaging Centre, Kings College London, London, 4 Department of Haematology, Aberdeen Royal Infirmary, Aberdeen, 5 Mount Vernon Cancer Centre, Northwood, 6 Department of Haematology, UCL Cancer Institute, University College Centre of Imaging, London, 7 Department of Haematology, James Paget University Hospital, Great Yarmouth, 8 Department of Oncology, Addenbrookes Hospital, Cambridge University Teaching Hospitals, London, and 9 Department of Haematology, Norfolk and Norwich University Hospital, Norwich, UK Keywords: Hodgkin, Hodgkin lymphoma, treatment, chemotherapy, radiotherapy. The guideline group was selected to be representative of UK-based medical experts and patients representatives. MEDLINE and EMBASE were searched systematically for publications in English from January 1990 to June 2013 using the key words Hodgkin, Lymphoma, Treatment, Chemotherapy and Radiotherapy. References from relevant publications were also searched. The writing group produced the draft guideline, which was subsequently revised by consensus by members of the Haemato-Oncology Task Force of the British Committee for Standards in Haematology (BCSH). The guideline was then reviewed by a sounding board of approximately 50 UK haematologists and the BCSH and comments incorporated where appropriate. The GRADE system was used to quote levels and grades of evidence, details of which can be found in Appendix I. The objective of this guideline is to provide healthcare professionals with clear guidance on the management of patients with classical Hodgkin Lymphoma (HL). The guidance may not be appropriate for all patients with HL and in all cases individual patient circumstances may dictate an alternative approach. Key recommendations Pre-treatment evaluation Patients require pre-treatment blood evaluation including human immunodeficiency virus (HIV) serology (1A). Staging with contrast enhanced computerized tomography (CT) of the neck to pelvis is required (1A), although Correspondence: Department of Haematology, Addenbrookes Hospital, Cambridge University Teaching Hospitals, Cambridge, UK. s: bcsh@b-s-h.org.uk; george.follows@addenbrookes.nhs.uk positron-emission tomography (PET)/CT is preferable if clinically feasible (1B). Early stage patients should be classified as favourable or unfavourable (1A). Advanced stage patients should be assessed to define the Hasenclever/International Prognostic Score (IPS) (1A). For male patients, pre-treatment semen cryopreservation should be offered where possible (1A). For female patients, pre-treatment review of options with a fertility specialist should be considered (1A). Management of early stage disease Prognostic factors should be determined to allocate patients to favourable and unfavourable sub-groups (1A). Standard of care for patients with favourable early stage HL is 2 3 ABVD (doxorubicin, bleomycin, vinblastine, dacarbazine) and 20 Gy radiotherapy (RT) (1A). Standard of care for unfavourable early stage HL is 4 3 ABVD and 30 Gy RT (1A). A treatment option for unfavourable early stage HL is 23 escalated BEACOPP (bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, prednisone) ABVD and 30 Gy RT (1A). The decision to omit RT from the management of stage IA/IIA non-bulky patients should involve discussion with a radiation oncologist (1B) and patients choosing to omit RT need to be aware of the balance of risks between RT and additional cycles of chemotherapy (1B). RT should not normally be omitted in patients presenting with bulky disease (1B). Early stage patients treated without RT should receive at least 3 3 ABVD (1B). Patients receiving bleomycin should be assessed carefully for signs and symptoms of pulmonary toxicity before each dose. A history of new or worsening dyspnoea or ª 2014 John Wiley & Sons Ltd, British Journal of Haematology doi: /bjh.12878

2 pulmonary crackles should lead to stopping of bleomycin until an alternative cause is identified (1B). Management of advanced stage disease Patients aged years old with advanced stage HL should receive either 6 8 cycles of ABVD or six cycles of escalated BEACOPP (1A). The choice between ABVD and escalated BEACOPP will depend on a range of factors, particularly the patient s opinion on the toxicity/efficacy balance between the regimens (2B). Patients with a higher IPS are at higher risk of relapse, potentially supporting the use of escalated BEACOPP in this higher risk group, although there are no prospective trial data to support a specific IPS cut-off at which escalated BEACOPP may be advantageous (2B). Patients treated with escalated BEACOPP who achieve an end-of-treatment PET-negative remission do not require consolidation RT to residual tissue (1A). Patients treated with ABVD should be considered for RT to sites of original bulk or residual tissue >15 cm. It remains unclear whether RT can be safely omitted in ABVD patients who have residual tissue >15 cm on CT that is PET-negative (1A). Interim PET (ipet2) is highly predictive of outcome in patients treated with ABVD (1A). It remains unclear how ipet2-positive patients are optimally managed in routine practice. Accepting the limitations of small, published datasets, treatment intensification to escalated BEACOPP RT appears reasonable (2B). Patients who remain PET-positive on completion of therapy require biopsy assessment or close clinical/ radiological surveillance for early progression (1B). Patients who develop progressive disease on therapy should be considered for treatment intensification with transplantation (see separate guidelines) (1A). Management of HL in pregnancy Patients should be closely co-managed with a speciald obstetric/fetal medicine unit (1B). Staging investigations and response evaluation should be tailored to the clinical presentation with radiology input to minimize fetal radiation exposure (1C). Delaying commencement of chemotherapy until postdelivery would not be standard practice and should be done with caution (1C). ABVD is the regimen of choice unless specifically contraindicated (1B). Wherever possible, RT should be delayed until postdelivery (1B). Management of HL in elderly patients Elderly patients should be formally assessed for fitness to receive combination chemotherapy with a co-morbidity assessment tool, which should distinguish frail from non-frail patients (2B). Patients considered frail should not usually be offered conventional combination chemotherapy (2B). Non-frail patients should be offered combination chemotherapy and RT with the aim of achieving complete remission (CR), which is associated with better survival (1B). Older patients receiving bleomycin must be followed very closely for symptoms and signs of bleomycin lung toxicity (1A). Guidance on therapy choice for non-frail patients is hampered by the lack of randomized trial data. Treatment with VEPEMB (vinblastine, cyclophosphamide, prednisolone, procarbazine, etoposide, mitoxantrone) or COPP (cyclophosphamide, vincristin, procarbazine, prednisone)/abvd appears to have lower treatmentrelated mortality than ABVD or BEACOPP (2B). PET/CT in HL PET/CT should be reported by PET/CT imaging specialists (1C). As pre-treatment staging with PET/CT will upstage a minority of patients and aid the interpretation of subsequent PET/CT, it is recommended when clinically feasible (1B). PET/CT response should be reported according to Deauville criteria (2B). By Deauville criteria, a score of 1 or 2 should be considered negative and 4 or 5 considered positive. Deauville score 3 should be interpreted according to the clinical context but in many HL patients indicates a good prognosis with standard treatment (1B). Biopsy is advised prior to second-line therapy to confirm residual disease with a score of 4 or 5 where possible to exclude false-positive uptake with fluorodeoxyglucose (FDG) (1B). The optimal management of ipet2-positive patients remains uncertain. Therefore, at this time, ipet2 remains desirable for ABVD-treated patients but cannot be mandated as a standard of care (2B). If a pre-treatment decision has been made to treat an early stage patient with RT following ABVD, then there is no clear role for interim PET/CT (1A). End-of-treatment PET/CT is recommended for all patients who have not achieved an interim PET-negative remission as this may directly affect RT planning, biopsy considerations and follow-up strategy (1B). 2 ª 2014 John Wiley & Sons Ltd, British Journal of Haematology

3 Radiotherapy strategies in HL The evidence for the role of RT in HL is based on involved field RT (IFRT) (1A). Reduced volume approaches, involved node (INRT) or involved site (ISRT) are under evaluation in current protocols (2B). The dose for favourable early stage disease should be 20 Gy, and 30 Gy for all other patients (1A). Follow-up, late effects and survivorship Patients are usually followed with intermittent outpatient clinical review for 2 5 years following first line therapy (2C). There is no proven role for routine surveillance CT or PET/CT imaging in patients who are otherwise well following first line therapy (2B). HL patients should be made aware that they are at an increased lifetime risk of second neoplasms, cardiovascular and pulmonary disease and infertility (1A). Apart from the current breast cancer-screening programme, there are no national cancer screening programmes tailored for HL survivors. Women treated with mediastinal RT before the age of 35 years should be offered entry into the breast cancer National Notification Risk Assessment and Screening programme (NRASP) (1A). Regular lifestyle advice should be offered to reduce secondary neoplasms and cardiovascular risk. There should be complete avoidance of smoking and careful management of cardiovascular risks, such as hypertension, diabetes mellitus and hyperlipidaemia (1B). Patients who have had RT to the neck and upper mediastinum should have regular thyroid function checks. Hypothyroidism can occur up to 30 years after RT (1A). Patients should receive irradiated blood products for life (1B). Background The annual incidence of Hodgkin Lymphoma (HL) in the UK is 27/ with approximately 1700 new cases per annum with a slight male predominance (Cancer Research UK, 2010). There is a peak in incidence in young adults aged years with a further peak observed >70 years of age. The incidence is currently stable (Morton et al, 2006; Cancer Research UK, 2010; Shenoy et al, 2011). Hodgkin Lymphoma is characterized by the presence of Hodgkin and Reed-Sternberg (HRS) cells within a cellular infiltrate of non-malignant inflammatory cells that make up the majority of the tumour tissue (Swerdlow et al, 2008). HRS cells have only recently been identified as clonal B cells that lack typical B-cell surface antigens. B cells failing to express surface immunoglobulin usually undergo apoptosis but HRS cells evade cell death through a number of mechanisms, including incorporation of Epstein-Barr virus (EBV) latent membrane proteins (LMP1 and 2), constitutive activation of the transcription factor nuclear factor jb (NFjB), and interaction with components of the microenvironment (Swerdlow et al, 2008; Steidl et al, 2011; Kuppers et al, 2012). Hodgkin Lymphoma is classified as either nodular lymphocyte predominant (NLPHL) or classical HL. There are four sub-types of classical HL: nodular sclerosis, mixed cellularity, lymphocyte-rich and lymphocyte-depleted; there is no difference in the prognosis or management of the different sub-types of classical HL. In Europe and North America, nodular sclerosis classical HL accounts for 70% of all classical HL. Lymphocyte-depleted classical HL is more prevalent in immunocompromised patients and is seen more commonly in developing countries, where it has a strong association with EBV infection. NLPHL is distinct histologically and HRS cells are not present, it has a risk of transformation to high grade non-hodgkin lymphoma and is managed differently from classical HL (Swerdlow et al, 2008). Clinical presentation of classical HL is usually with painless lymphadenopathy, which is most commonly cervical or supraclavicular. Mediastinal disease is identified in 80% of patients and is more common in nodular sclerosing HL, whilst peripheral or sub-diaphragmatic lymphadenopathy is more common in mixed-cellularity classical HL. Bone marrow involvement is detected in only 5 8% of patients with conventional staging (Mauch et al, 1993; Levis et al, 2004a; Swerdlow et al, 2008), but in up to 18% with positron emission tomography (PET)/ computerized tomography (CT) staging (El-Galaly et al, 2012). Systemic symptoms of drenching night sweats, unexplained fever >38 C, and weight loss of >10% over 6 months are termed B symptoms and are identified in approximately 25% of patients. Pretreatment evaluation Blood evaluation should include full blood count, erythrocyte sedimentation rate, renal function, liver function, bone profile, lactate dehydrogenase and testing for human immunodeficiency virus (HIV). Patients should be staged with a contrast-enhanced CT scan covering the neck, chest, abdomen and pelvis. An initial PET/ CT scan is highly recommended as this provides a baseline for interpretation of subsequent scans and, in a minority of cases, it can upstage patients and alter the planned therapy. It is appreciated that a minority of patients may present with very advanced disease and if obtaining a PET/CT scan would result in a treatment delay this may not be clinically appropriate. It was common practice to limit bone marrow evaluation to patients with advanced stage disease or B symptoms, however, it is now generally accepted that PET/CT can accurately detect marrow involvement and evaluation by biopsy is often unnecessary (El-Galaly et al, 2012). Clinicians should be ª 2014 John Wiley & Sons Ltd, British Journal of Haematology 3

4 aware that diffuse bone marrow uptake on PET may just reflect a reactive process. Patients with early stage disease should be categorized as having favourable or unfavourable characteristics. Patients with advanced stage disease should have their Hasenclever/ International Prognostic Score (IPS) determined (Hasenclever & Diehl, 1998). Consideration should be given to fertility preservation and semen cryopreservation should be offered routinely before therapy with combination chemotherapy. There is increasing evidence for the effectiveness of oocyte preservation as a fertility-sparing strategy and referral to a fertility specialist should be considered, if treatment delays are acceptable. This is especially important if the plan is to administer escalated BEACOPP (bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, prednisone). ABVD (doxorubicin, bleomycin, vinblastine, dacarbazine) is generally considered to be fertility-sparing, but some patients may need to receive salvage chemotherapy and stem cell transplantation, which can result in a reduction of fertility. The role for prophylactic use of gonadotropin-releasing hormone (GnRH) analogues to preserve female fertility remains uncertain (Behringer et al, 2012; Wong et al, 2013). For women with a stable partner, and in whom a delay of treatment is possible, in vitro fertilization for embryo cryopreservation may be appropriate. However, this may not be widely available at short notice. Ovarian tissue cryopreservation remains experimental and so far has resulted in only a small number of pregnancies and births (Loren et al, 2013). Generally, fertility preservation techniques can only be discussed on a case-by-case basis, involving fertility experts and oncologists who can judge the appropriateness of the techniques and the risks of delaying treatment. Key recommendations for pre-treatment evaluation Patients require pre-treatment blood evaluation including HIV serology (1A). Staging with contrast enhanced CT of the neck to pelvis is required (1A), although PET/CT is preferable if clinically feasible (1B). Early stage patients should be classified as favourable or unfavourable (1A). Advanced stage patients should be assessed to define the Hasenclever/IPS (1A). For male patients, pre-treatment semen cryopreservation should be offered where possible (1A). For female patients, pre-treatment review of options with a fertility specialist should be considered (1A). Management of early stage disease Randomized trials have shown that combined modality treatment with chemotherapy and radiotherapy (RT) results in superior tumour control compared with RT alone (Noordijk et al, 2006; Engert et al, 2007; Ferme et al, 2007). Clinical trials in patients with Stage I II disease have increasingly evaluated treatment reduction as a strategy to reduce late morbidity and mortality. These trials have defined and refined prognostic factors for favourable and unfavourable prognosis Stage I II HL. Traditionally in the UK, patients with early stage HL with B symptoms or bulk disease have been managed with protocols for advanced stage disease. However, long-term follow-up of large trial data sets, such as the German Hodgkin Study Group (GHSG) HD10 trial, have confirmed that these patients generally have excellent outcomes when treated with early stage unfavourable risk protocols. Definition of large mediastinal mass /bulky disease varies slightly between study groups: The European Organization for the Research and Treatment of Cancer (EORTC) defines bulk as a mediastinal thoracic ratio >035 at T5/6. The UK National Cancer Research Institute (NCRI) and GHSG define a mediastinal mass ratio >033 as large, while the US National Comprehensive Cancer Network (NCCN) and National Cancer Institute of Canada (NCIC) define bulky as a mediastinal mass ratio>033 or any mass with maximal diameter >10 cm. Prognostic factors for the EORTC and GHSG are listed in Tables I and II. Favourable early stage disease Recent trials have focussed on abbreviating chemotherapy and RT dosages, and assessing RT-free treatment strategies. The GHSG HD10 trial randomized 1190 patients into four treatment arms, which included two cycles of ABVD plus 30 Gy RT, two cycles of ABVD plus 20 Gy of RT, four cycles of ABVD plus 30 Gy of RT and four cycles of ABVD plus 20 Gy of RT. With a 76-year median follow up, no difference was observed in freedom from progression (97%) or overall survival (OS, 98%) between the four groups (Engert et al, 2010). Thus, the least toxic approach consisting of two cycles of ABVD followed by 20 Gy RT appears to be sufficient in favourable Stage I II HL. The trial with the longest published median follow-up (113 years) in early stage HL is the IA/IIA non-bulky NCIC and Eastern Cooperative Oncology Group (ECOG) trial comparing ABVD alone (4 6 cycles) with treatment includ- Table I. Favourable prognosis Stage I II Hodgkin Lymphoma. EORTC GHSG No large mediastinal adenopathy No large mediastinal adenopathy ESR <50 without B symptoms ESR <50 without B symptoms ESR <30 with B symptoms ESR <30 with B symptoms Age 50 years No extranodal disease 1 3 lymph node sites involved 1 2 lymph node sites involved EORTC, European Organization for the Research and Treatment of Cancer; GHSG, German Hodgkin Study Group; ESR, erythrocyte sedimentation rate. 4 ª 2014 John Wiley & Sons Ltd, British Journal of Haematology

5 Table II. Unfavourable prognosis Stage I II Hodgkin lymphoma. EORTC presence of one or more of the following Large mediastinal adenopathy ESR 50 without B symptoms ESR 30 with B symptoms Age >50 years 4 lymph node sites involved GHSG presence of one or more of the following Large mediastinal adenopathy ESR 50 without B symptoms ESR 30 with B symptoms Extranodal disease 3 lymph node sites involved EORTC, European Organization for the Research and Treatment of Cancer; GHSG, German Hodgkin Study Group; ESR, erythrocyte sedimentation rate. ing subtotal nodal irradiation (STNI) (Meyer et al, 2012). From separate sequential analysis of this trial (Meyer et al, 2005, 2012), it appears that omitting RT increases the risk of early relapse, but with longer follow-up does not appear to compromise OS. It is difficult to interpret the comparative aspects of this trial as STNI is no longer used in routine practice, and there were a small number of unusual events in the STNI arm that complicated the interpretation of OS of the RT-arm of the study. However, this trial has shown that treating IA/IIA non-bulky HL patients with ABVD alone delivers long-term OS of 94%. Data from the UK NCRI trial (RAPID) and the EORTC H10 trial were presented at the annual American Society of Hematology meeting in 2012 (Andre et al, 2012; Radford et al, 2012). In both trials, patients who achieved an early PET-negative remission had a better progression-free survival (PFS) than interim PET-positive patients. Both trials randomized the patients who achieved an early PET-negative remission with ABVD chemotherapy to receive or omit RT. With both trials, there was an early PFS advantage with the inclusion of RT, but no OS advantage had been shown within the limited follow-up periods of both trials. Of note, patients with mediastinal bulk disease were excluded from the RAPID trial. In the RAPID study, patients who were PET-positive after three ABVD cycles received 1 9 additional ABVD and RT. This achieved a 3-year PFS of 859% and OS of 939% (Radford et al, 2012). With the current available data, the standard of care for treatment of patients with early stage favourable HL (as defined by GHSG criteria) therefore remains ABVD RT. However, as there are longer term risks for patients treated with RT, it is recognized that, in some circumstances, clinicians and patients may prefer to treat without RT, particularly as the majority of IA/IIA non-bulky patients will be cured with chemotherapy alone. Indeed, a cross-trial comparative analysis between the German HD10/11 and NCIC HD6, suggested patients who achieve a CT complete remission (CR) after 2 9 ABVD, and then complete a total 4 9 ABVD with no RT, have the same excellent outcome as 2 9 ABVD + RT (Hay et al, 2012). However, the decision to treat early stage good risk patients without RT should involve a radiation oncologist and the patient must be made aware that a number of trials indicate that removing RT probably increases the risk of early relapse by approximately 3 7% (Meyer et al, 2005, 2012; Andre et al, 2012; Radford et al, 2012). If patients are treated without RT, then choosing the optimum number of cycles of ABVD is difficult. The NCIC/ ECOG trial, using 4 6 cycles of ABVD, has the longest followup, but initial data from 3 9 ABVD for PET-negative patients in the RAPID trial are encouraging. A balance therefore needs to be struck between long-term toxicity of RT and the toxicity of the additional ABVD cycles (Swerdlow et al, 2011). Apart from abstract presentations of the UK NCRI trial (RAPID) and the EORTC H10 trial (Andre et al, 2012; Radford et al, 2012), larger randomized trials using PET/CT to guide decision-making in early stage HL have yet to be formally reported. Consensus opinion from all trials groups is that early stage patients with persisting PET-avid disease after chemotherapy should receive RT. At this stage, it remains uncertain whether achieving an interim PET-negative remission after two cycles of ABVD is as predictive for long-term PFS and OS as achieving a CR by established CT criteria. Unfavourable early stage disease In the HD11 trial, the GHSG randomly assigned 1395 patients with early unfavourable HL to four cycles of ABVD plus 30 Gy RT, four cycles of ABVD plus 20 Gy RT, four cycles of BEA- COPP plus 30 Gy RT and four cycles of BEACOPP plus 20 Gy RT (Eich et al, 2010). With 68 years median follow-up, no difference was observed in OS (93 96%) for all four groups. In the arms of the study with 30 Gy of RT, there was no difference in freedom from treatment failure (FFTF) between BEACOPP and ABVD (P = 065), but a significant difference in favour of BEA- COPP was seen for FFTF when 20 Gy RT was used (P = 002) (Eich et al, 2010). The German Hodgkin study Group HD14 Trial randomly assigned early unfavourable HL patients to either four cycles of ABVD or an intensified treatment of two cycles of escalated BEACOPP followed by two cycles of ABVD (2 + 2) (von Tresckow et al, 2012). Chemotherapy was followed by 30 Gy RT in both arms. At 5 years follow-up, the dose-intensified regime resulted in better tumour control with a 62% improvement in PFS compared with standard treatment with four cycles of ABVD. However, acute toxicities were significantly higher in the arm and with no difference in OS, this regimen would be considered a treatment option rather than a standard of care for the majority of patients. Currently, four cycles of ABVD followed by 30 Gy RT is widely considered the standard of care for unfavourable early stage HL. There is no evidence to support the removal of RT from patients who present with bulky disease, and of note, such patients were excluded from the RAPID and NCIC/ ECOG 6 trials. Infradiaphragmatic early stage disease The incidence of infradiaphragmatic early stage HL is very low and accounts for 4 13% of cases of Stage I II disease ª 2014 John Wiley & Sons Ltd, British Journal of Haematology 5

6 (Vassilakopoulos et al, 2006). When results are adjusted for risk factors they appear to have a similar prognosis to patients with supradiaphragmatic disease. Older age, clinical Stage II (borderline), involvement of 3 sites and higher IPS is more frequent in patients with infradiaphragmatic disease and the previously reported poorer outcome may be explained by the unfavourable profile of the patients. Although this group of patients are under-represented in clinical trials, there is currently no evidence to suggest that they should be treated differently from their supradiaphragmatic counterparts (Darabi et al, 2005; Vassilakopoulos et al, 2006). KEY recommendations for management of early stage disease Prognostic factors should be determined to allocate patients to favourable and unfavourable sub-groups (1A). Standard of care for patients with favourable early stage HL is 2 3 ABVD and 20 Gy RT (1A). Standard of care for unfavourable early stage HL is 4 3 ABVD and 30 Gy RT (1A). A treatment option for unfavourable early stage HL is 23 escalated BEACOPP ABVD and 30 Gy RT (1A). The decision to omit RT from the management of stage IA/IIA non-bulky patients should involve discussion with a radiation oncologist (1B) and patients choosing to omit RT need to be aware of the balance of risks between RT and additional cycles of chemotherapy (1B). RT should not normally be omitted in patients presenting with bulky disease (1B). Early stage patients treated without RT should receive at least 3 3 ABVD (1B). Patients receiving bleomycin should be assessed carefully for signs and symptoms of pulmonary toxicity before each dose. A history of new or worsening dyspnoea or pulmonary crackles should lead to stopping of bleomycin until an alternative cause is identified (1B). Advanced stage disease Standard treatment for patients with stage III/IV HL is combination chemotherapy. In the UK, patients with early stage disease but B symptoms or bulky disease have usually been managed with advanced HL protocols, although this is not universal international practice (Townsend & Linch, 2012). There is significant international variation in the use of RT to sites of initial bulk disease or residual masses following chemotherapy. Over the past three decades of comparative randomized clinical trials in HL, ABVD and escalated BEACOPP have become established international standards of care for advanced HL [reviewed in (Townsend & Linch, 2012)]. ABVD has better outcomes in terms of efficacy and/or toxicity than MOPP (mechlorethamine, vinristine, procarbazine, predisone), MOPP ABV (MOPP, doxorubicin, bleomycin, vinblastine) hybrid and ChlVPP/PABLOE (chlorambucil, vinblastine, procarbazine, prednisone/doxorubicin, bleomycin, vincristine and etoposide), and similar efficacy, but lower toxicity, than the Stanford V regimen (mechlorethamine, doxorubicin, vinblastine, prednisone, vincristine, bleomycin, etoposide) (Canellos et al, 1992; Duggan et al, 2003; Johnson et al, 2005; Hoskin et al, 2009). From a number of advanced HL trials, failure-free/pfs with ABVD is around 73 78% with OS in the range of 82 90% (Duggan et al, 2003; Johnson et al, 2005; Federico et al, 2009; Hoskin et al, 2009; Viviani et al, 2011). ABVD should be delivered on schedule with infusions given every 14 d irrespective of neutrophil count. Granulocyte colony-stimulating factor (G-CSF) is only required for patients with infectious complications (Evens et al, 2007; Nangalia et al, 2008). Different HL study groups have followed different strategies for RT following chemotherapy. Patients with bulky advanced stage disease who achieve a CT CR after MOPP/ AVD can avoid RT in contrast with patients in partial remission (PR), who did derive benefit from RT (Aleman et al, 2003). It remains unclear as to whether RT to sites of initial bulk disease or residual tissue can be safely omitted in patients treated with ABVD. In the LY09 trial, 43% of patients received RT following ABVD or ChlVPP/PABLOE and, with a median follow-up of 69 years, these patients had a PFS and OS advantage compared with patients treated without RT (Johnson et al, 2005). Of note, the ABVD trials with the highest OS rates have used RT for a large proportion of patients (Duggan et al, 2003; Johnson et al, 2005; Hoskin et al, 2009). Escalated BEACOPP was been developed by the German HL study group (GHLSG). In comparison with COPP ABVD and standard dose BEACOPP, escalated BEACOPP has demonstrated improved PFS and OS (Engert et al, 2009). Data from 2182 patients randomized from 408 centres in the GHLSG HD15 trial have shown a 5-year failure-free and OS of 89% and 95% respectively, for patients receiving six cycles of escalated BEACOPP (Engert et al, 2012). Patients in this trial who achieved a PET-negative remission (but PR by conventional CT) were not treated with RT consolidation, in contrast with the HD9 trial. The number of patients receiving RT fell from 71% in HD9 to 11% in HD15, with no apparent loss in treatment efficacy (Engert et al, 2012). Although this is not prospective randomized data, it is reasonable to conclude that patients who achieve PET-negative remission following escalated BEACOPP therapy do not require RT to sites of residual tissue (Engert et al, 2012). From HD15, the outcomes for patients treated with escalated BEACOPP 9 6 or BEACOPP were similar and clinicians/patients may prefer to be treated on the latter regimen if appropriate. Escalated BEACOPP-based strategies have also delivered impressive results in high-risk childhood/ 6 ª 2014 John Wiley & Sons Ltd, British Journal of Haematology

7 adolescent HL [5-year event-free survival (EFS)/OS: 94%/ 97%] (Kelly et al, 2011), but have shown unacceptable treatment-related toxicity in patients >60 years old (Engert et al, 2009) and the GHLSG no longer recommends this protocol for patients over 60 years of age. Data directly comparing ABVD with escalated BEACOPP are limited. The EORTC trial comparing ABVD with BEACOPP (49 escalated, 49 standard) for higher risk HL patients has been presented in abstract form only (Carde et al, 2012). A PFS advantage was found for escalated BEA- COPP, but no survival advantage has yet been demonstrated with a median follow-up of 38 years. A smaller Italian trial randomized patients to either ABVD or escalated BEACOPP followed by treatment intensification with RT (to sites of initial bulk and residual tissue) then salvage chemotherapy/ autologous transplant for all patients failing to achieve a CR (Viviani et al, 2011). Initial treatment with escalated BEA- COPP resulted in a 7-year freedom-from-first progression advantage compared with ABVD (85% vs. 73%; P = 0004), but the 7-year OS was no different, at 89% and 84%, respectively (P = 032). Considering the first line intensification strategy given to approximately a quarter of the patients and significant criticisms raised with regards to this trial (Borchmann et al, 2011; Corazzelli et al, 2011; Tam et al, 2011), it is difficult to interpret the results in the context of standard UK practice. However, the data support the consensus view that, compared with ABVD, escalated BEA- COPP provides a PFS advantage for patients, while adding to the intense debate as to whether first line escalated BEA- COPP provides a true survival advantage for advanced HL patients. ABVD has a better toxicity profile than escalated BEA- COPP in terms of direct side effects of chemotherapy, infectious complications, blood product requirements, fertility preservation and secondary myelodysplastic syndrome (MDS)/acute myeloid leukaemia (AML) (Sieniawski et al, 2008; Engert et al, 2009). The HD15 data confirm that the short- and long-term toxicity profile of escalated BEACOPP is improved significantly by limiting treatment to 6 cycles (Engert et al, 2012). The cumulative total doses of doxorubicin and bleomycin are lower for patients treated with 69 escalated BEACOPP compared with 6 9 ABVD (210 mg/m 2 and iu/m 2 compared with 300 mg/m 2 and iu/m 2, respectively). With a median follow-up of 4 years, the cumulative incidence of MDS/AML is 03% in patients treated with 6 cycles of escalated BEACOPP (Engert et al, 2012). Interestingly the EORTC trial found no difference in second cancer/mds rates between ABVD and escalated BEA- COPP (Carde et al, 2012). Rates of female infertility are strongly influenced by the chemotherapy choice and age at treatment (Behringer et al, 2005, 2013; De Bruin et al, 2008). With 4 cycles of ABVD within the GHLSG HD14 trial, 100% of women aged <30 years and 94% aged >30 years regained regular menstrual cycles following chemotherapy. Data from patients treated with 6 8 cycles of ABVD are more limited, but there does not appear to be additional loss of fertility with the additional cycles of treatment (Hodgson et al, 2007a; De Bruin et al, 2008). With 6 8 cycles of escalated BEACOPP within HD15, the corresponding figures were 82% of women aged <30 regained regular menstrual cycles compared with 45% aged >30 (Behringer et al, 2005, 2013; De Bruin et al, 2008). A positive PET/CT scan after two cycles of ABVD, interim PET (ipet), has been shown to be strongly predictive of treatment failure for patients who continue with ABVD therapy (Gallamini et al, 2007). A number of recently closed and ongoing therapeutic trials are assessing whether, in comparison with historical controls, treatment intensification for ipet2 positive patients with escalated BEACOPP can improve patient outcomes (Gallamini et al, 2011, 2012). The UK RATHL trial ( uk/search/studydetail.aspx?studyid=4488) chose to intensify therapy with either 49 escalated BEACOPP or 4 9 BEA- COPP-14 with no RT requirement, while the Italian approach has been to use 8 cycles of BEACOPP (49 escalated and 49 standard) and RT to sites of initial bulk disease. Initial reports from the Italian group appear encouraging (Gallamini et al, 2011, 2012). There is no international consensus on whether patients are best served by starting therapy with either ABVD or escalated BEACOPP and personal priorities for each patient will clearly influence this decision. There are no prospective data to support a specific IPS cut-off when deciding between escalated BEACOPP and ABVD. Patients opting for ABVD will have clear benefits in terms of toxicity, but will have a higher risk of relapse and need for salvage therapy/autologous transplant. With ABVD-treated patients, there is also uncertainty with regards to the need for RT to sites of initial bulky disease/residual tissue and lack of clarity as to the role of ipet2 when treated outside a clinical trial. Patients opting for escalated BEACOPP have clear benefits with regards to EFS and the clarity that RT will be needed in only 10% of patients. They will, however, potentially experience more treatmentrelated toxicity. There are no data on the use of escalated BEACOPP in HIV-related HL, where standard management remains ABVD combined with anti-retroviral therapy (Montoto et al, 2012). Key recommendations for management of advanced stage disease Patients aged years with advanced stage HL should receive either 6 8 cycles of ABVD or 6 cycles of escalated BEACOPP (1A). The choice between ABVD and escalated BEACOPP will depend on a range of factors, particularly the patient s opinion on the toxicity/efficacy balance between the regimens (2B). ª 2014 John Wiley & Sons Ltd, British Journal of Haematology 7

8 Patients with a higher IPS are at higher risk of relapse, potentially supporting the use of escalated BEACOPP in this higher risk group, although there are no prospective trial data to support a specific IPS cut-off at which escalated BEACOPP may be advantageous (2B). Patients treated with escalated BEACOPP who achieve an end of treatment PET-negative remission do not require consolidation RT to residual tissue (1A). Patients treated with ABVD should be considered for RT to sites of original bulk or residual tissue >15 cm. It remains unclear whether RT can be safely omitted in ABVD patients who have residual tissue >15 cm on CT that is PET-negative (1A). Interim PET (ipet2) is highly predictive of outcome in patients treated with ABVD (1A). It remains unclear how ipet2-positive patients are optimally managed in routine practice. Accepting the limitations of small, published datasets, treatment intensification to escalated BEACOPP RT appears reasonable (2B). Patients who remain PET-positive on completion of therapy require biopsy assessment or close clinical/ radiological surveillance for early progression (1B). Patients who develop progressive disease on therapy should be considered for treatment intensification with transplantation (see separate guidelines) (1A). Management of HL in pregnancy There are no prospective trials for the management of HL in pregnancy, but a number of published case reports, case series and case cohorts were reviewed by Bachanova and Connors (2008). The priority must be the health of the mother and, ideally, management should be in conjunction with an obstetrician experienced in high-risk pregnancy. While termination of pregnancy may be the most appropriate course of action for certain patients, for many cases, the fetal and maternal outcomes following HL treatment in pregnancy appear excellent (Evens et al, 2011). Staging and response assessment with magnetic resonance imaging scans and ultrasound may avoid the need for radiation-based imaging and, in certain cases, delaying therapy until post-partum may be possible, although this must be done with caution. If therapy is required in pregnancy, the general consensus is that ABVD is the regimen of choice if multi-agent chemotherapy is to be used (Bachanova & Connors, 2008). Although ABVD has been used to treat patients in all 3 trimesters (Anselmo et al, 1999; Cardonick & Iacobucci, 2004), the potential risk to fetal development from chemotherapy is likely to be higher in the first trimester and most clinicians would try and avoid exposure to chemotherapy at this time. Wherever possible, RT should be delayed until post-delivery. Key recommendations for HL in pregnancy Patients should be closely co-managed with a specialized obstetric/fetal medicine unit (1B). Staging investigations and response evaluation should be tailored to the clinical presentation with radiology input to minimize fetal radiation exposure (1C). Delaying commencement of chemotherapy until postdelivery would not be standard practice and should be done with caution (1C). ABVD is the regimen of choice unless specifically contraindicated (1B). Wherever possible, RT should be delayed until post delivery (1B). Management of HL in elderly patients Population based data suggest that the over-60s age group account for approximately 20% of cases of HL (Stark et al, 2002) and a number of studies have suggested that their outcome is consistently less good than that of younger patients, especially for those with advanced stage disease (Evens et al, 2008, 2012, 2013; Proctor et al, 2009). Older patients are more likely to die from non-hl causes, including bleomycin lung toxicity, which was reported in 24% of patients over 60 years old treated within the North American Intergroup trial E2496 (Evens et al, 2013). Therapy-related toxicity has, therefore, made delivering the gold standard chemotherapies, ABVD and BEACOPP, challenging, especially in the over-70s and the major randomized trials contain only small numbers of elderly patients. Following a number of workshops at international lymphoma meetings in the early 2000s it was decided to pursue a series of non-randomized phase II studies of alternative therapies in elderly HL. The GHSG have developed phase 2 trials with BACOPP (modified and dose-reduced BEACOPP) and PVAG (prednisolone, vincristine, doxorubicin, gemcitabine) (Halbsguth et al, 2010; Boll et al, 2011). BACOPP (bleomycin, adriamycin, cyclophosphamide, vincristine, procarbazine, prednisone) was used to treat 65 patients aged years, producing a CR rate of 85% but with a 12% treatment-related mortality and 33 month OS of 70%. With PVAG, 59 patients aged years were treated for mainly advanced stage disease, producing a 3-year PFS and OS of 58% and 66%. The GHSG have also analysed the outcomes for early stage elderly patients treated with ABVD within the HD10 and HD11 trials (Boll et al, 2013). They observed excellent response rates (CR of 89%), but a treatment-related mortality of 5%. With 92 months median follow-up, the 5-year PFS was estimated at 75%. With the UK Shield registration study, the CR rate for early stage patients treated with two or more cycles of ABVD and involved field RT (IFRT) was 62% but, for advanced stage ABVD patients, the CR rate was only 46%, and there were 18% treatment-related deaths (Proctor et al, 2012). Of note, within this registration study, no patients 8 ª 2014 John Wiley & Sons Ltd, British Journal of Haematology

9 who were classified as frail based on a scoring system, completed treatment or achieved a CR. The high treatmentrelated death rate with ABVD is similar to that described in the ABVD versus Stamford V trial for older patients (Evens et al, 2013). In the German HD9 elderly analysis, acute toxicity led to 21% treatment-related deaths after BEACOPP compared with 8% for COPP/ABVD (Ballova et al, 2005). UK and Italian groups have both explored a treatment strategy with VEPEMB (vinblastine, cyclophosphamide, prednisolone, procarbazine, etoposide, mitoxantrone). In total, 103 patients were treated with VEPEMB as part of the Shield programme (Proctor et al, 2012), including 31 early stage patients (VEPEMB 9 3 followed by involved field RT) with a CR rate of 74% and 3-year PFS and OS of 74% and 81%, respectively. For 72 advanced stage patients treated with VEPEMB 9 6a CR rate of 61% was achieved with 3-year PFS and OS of 58% and 66%, respectively. However, patients classified as frail were excluded from the UK VEPEMB trial and this probably improved the results. Nevertheless, similar encouraging data with VEPEMB have been reported from an Italian trial (Levis et al, 2004b). It also remains common practice to treat less fit elderly patients with older non-anthracycline containing regimens, such as ChlVPP, although published outcomes for this regimen in the elderly are generally poor (The International ChlVPP Treatment Group, 1995). Key recommendations for management of elderly patients Elderly patients should be formally assessed for fitness to receive combination chemotherapy with a co-morbidity assessment tool, which should identify frail from non-frail patients (2B). Patients considered frail should not usually be offered conventional combination chemotherapy (2B). Non-frail patients should be offered combination chemotherapy and RT with the aim of achieving CR, which is associated with better survival (1B). Older patients receiving bleomycin must be followed very closely for symptoms and signs of bleomycin lung toxicity (1A). Guidance on therapy choice for non-frail patients is hampered by the lack of randomized trial data. Treatment with VEPEMB or COPP/ABVD appears to have lower treatment-related mortality than ABVD or BEA- COPP (2B). The use of PET/CT in HL Hodgkin Lymphoma is 18 FFDG avid in % of cases (Weiler-Sagie et al, 2010). Staging patients with FDG PET-CT is more accurate than CT alone (Cheson et al, 2007) and the availability of a baseline scan increases the reliability of subsequent response assessment (Quarles van Ufford et al, 2010; Barrington et al, 2011). Contrast-enhanced CT may be done as part of the PET-CT examination or at a separate visit, depending on local imaging arrangements. The five-point scale (5-PS) also referred to as the Deauville criteria has been used for reporting in response guided trials and has published high inter-observer agreement (Barrington et al, 2010; Furth et al, 2011; Biggi et al, 2013) and improved predictive value (Le Roux et al, 2011) when compared with earlier International Harmonization criteria (Juweid et al, 2007). The response scan is compared with the baseline scan and scored according to the level of highest residual FDG uptake using the 5-PS as follows: Score 1: No uptake. Score 2: Uptake less than or equal to the mediastinum. Score 3: Uptake greater than the mediastinum but less than the liver. Score 4: Uptake moderately higher than the liver. Score 5: Uptake markedly higher than the liver. After chemotherapy, stimulation of normal bone marrow may result in diffusely increased uptake that is higher than normal liver. The uptake in sites of initial marrow involvement should then be compared with uptake within normal marrow to assess the presence/absence of residual disease. The 5-PS can be used to assess response during treatment (with an interim scan) and at the end of treatment. To avoid the risk of under-treatment, scores 1 and 2 were used to define a negative scan in the RAPID study (Radford et al, 2012). The mediastinum was also used as the threshold for satisfactory response in the HD15 study for patients treated with BEACOPP who subsequently did not receive RT (Engert et al, 2012). It is recommended therefore that a score of 1 or 2 is used to define a complete metabolic response (CMR) if omission of standard RT treatment is being considered in discussion with patients. To avoid the risk of over-treatment, scores 4 and 5 were used to define a positive scan in the NCRI RATHL study (Barrington et al, 2010) and in the Italian HD0607 study and scores 1, 2, 3 defined a negative scan ( with initial reports suggesting satisfactory responses for patients treated on trial (Gallamini et al, 2012; Johnson et al, 2013). Until the trials report in full, it seems prudent to recommend using score 4 or 5 to define an inadequate response to ABVD at interim, if therapy intensification to escalated BEACOPP is to be considered. Interim scans should ideally be performed as long after the last chemotherapy administration as possible to avoid potential false-positive uptake. At the end of treatment, score 4 or 5 probably represents an inadequate response, however, biopsy is recommended prior to second line treatment to exclude false-positive uptake reflecting treatment-related inflammation. Other causes of false-positive uptake, including infection, granulomatous disease and sarcoid-like reactions, are also recognized (Barrington & O Doherty, 2003). Patients with score 3, ª 2014 John Wiley & Sons Ltd, British Journal of Haematology 9

10 which was previously referred to as minimal residual uptake, (Hutchings et al, 2005) are likely have good outcomes with standard treatment (Hutchings et al, 2005; Biggi et al, 2013; Johnson et al, 2013), but until more information is available, de-escalation of therapy outside trials in these patients is not advised. Standardized methods for PET acquisition and quality control have been developed during the conduct of responseguided trials and are recommended for best clinical practice. The reliability of visual and quantitative assessments depends on correct patient preparation, image acquisition and quality assessment of imaging equipment. Therefore, if management changes are considered based on PET findings, guidelines for tumour imaging with FDG standards should be adhered to (Boellaard et al, 2010). Both interim and end-of-treatment scans should be performed as long after the last chemotherapy administration as possible, to avoid false-positive uptake due to inflammation induced by treatment (Spaepen et al, 2003; Boellaard et al, 2010). Therefore, interim scans should be performed as close to the start of the next cycle of chemotherapy as practical. At the end of treatment, a minimum of 10 d following chemotherapy, 2 weeks after granulocyte colonystimulating factor (G-CSF) treatment and 3 months after RT is recommended to elapse prior to scanning (Boellaard et al, 2010). Key recommendations for PET/CT in HL PET/CT should be reported by PET/CT imaging specialists (1C). As pre-treatment staging with PET/CT will upstage a minority of patients and aid the interpretation of subsequent PET/CT, it is recommended when clinically feasible (1A). PET/CT response should be reported according to Deauville criteria (2B). By Deauville criteria, a score of 1 or 2 should be considered negative and 4 or 5 considered positive. Deauville score 3 should be interpreted according to the clinical context but in many HL patients indicates a good prognosis with standard treatment (1B). Biopsy is advised prior to second-line therapy to confirm residual disease with a score of 4 or 5 where possible to exclude false-positive uptake with FDG (1B). The optimal management of ipet2 positive patients remains uncertain. Therefore, at this time, ipet2 remains desirable for ABVD-treated patients but cannot be mandated as a standard of care (2B). If a pre-treatment decision has been made to treat an early stage patient with RT following ABVD, then there is no clear role for interim PET/CT (1A). End-of-treatment PET/CT is recommended for all patients who have not achieved an interim PET negative remission as this may directly affect RT planning, biopsy considerations and follow-up strategy 1B. Radiotherapy strategies in HL As chemotherapy has become more effective, the role of RT in HL has diminished, but it is still an essential tool, maximizing local control, allowing fewer cycles of chemotherapy and avoiding intensive chemotherapy in relapse. The evidence base for RT in both early and advanced HL is based on IFRT. However, concerns regarding the late effects of radiation have led to new protocols using reduced volume treatment. An EORTC consensus led by Groupe d Etude des Lymphomes de l Adulte (GELA) is now in favour of involved node RT (INRT) (Girinsky et al, 2008). However, this depends on obtaining pre- and post-treatment PET-CT imaging of the patient in the treatment position (Girinsky et al, 2008; Specht et al, 2013). An alternative approach uses involved site RT (ISRT), which adds a 15 mm safety margin above and below the involved node (Hoskin et al, 2013; Specht et al, 2013). Larger volumes are still preferred in those rare cases where RT alone is recommended. The dose of RT is now well defined from the GHSG trials HD10 and HD11, which show that 20 Gy is sufficient for favourable early HL, and 30 Gy should be given to all other patients with early or advanced stage disease, where RT is indicated. Key recommendations for radiotherapy strategies in HL The evidence for the role of RT in HL is based on IFRT (1A). Reduced volume approaches, INRT or ISRT are under evaluation in current protocols (2B). The dose for favourable early stage disease should be 20 Gy, and 30 Gy for all other patients (1A). Follow-up, late effects and survivorship issues There is significant variation in follow-up practice with little evidence to support a particular strategy. Clinicians typically keep patients in follow-up clinics for 5 years before discharge with intermittent outpatient review, but follow-up frequency inevitably varies depending on patient requirements. Some patients/clinicians may prefer early discharge from formal follow-up after completion of first line therapy. Routine CT or PET/CT scans significantly increase radiation exposure and health care costs with no clear evidence of benefit for patients (Voss et al, 2012; Patel et al, 2013). As such, routine CT or PET/CT scanning for otherwise well patients is not normally required. Despite the high cure rates, long-term survivors of HL have increased mortality compared to the general population (Ng et al, 2002; Favier et al, 2009; Baxi & Matasar, 2010). During the first 5 10 years of follow-up, the lead cause of absolute excess risk (AER) of mortality remains HL, especially in those with unfavourable prognosis. However, the AER for all-cause 10 ª 2014 John Wiley & Sons Ltd, British Journal of Haematology